Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: pca and predict--confusion about what it does

From   Nick Cox <>
Subject   Re: st: pca and predict--confusion about what it does
Date   Sat, 20 Oct 2012 21:56:18 +0100

If you want another mean, all you need to do is add it. No option
really needed, therefore.


On Sat, Oct 20, 2012 at 8:17 PM, Israel Pearce <> wrote:
> Thank you. I can't seem to find any options on Stata that do not scale
> the PC's to have mean 0. Do you know of an option that could allow for
> this or is it not a feature in Stata?
> On Sat, Oct 20, 2012 at 11:02 AM, Nick Cox <> wrote:
>> The first predicted variable is the first PC, and so on. It's
>> conventional that PCs are scaled to have mean 0.
>> The absence of a response or outcome variable is irrelevant here. If
>> you're confused a good reason is because Stata is stretching the
>> meaning of -predict- here to include these constructed variables. (A
>> bad reason is not reading the documentation...)
>> Nick
>> On Sat, Oct 20, 2012 at 6:32 PM, Israel Pearce <> wrote:
>>> I am confused about the principal component scores one gets from pca
>>> postestimation. Let's say I wanted the first principal component
>>> scores from a set of explanatory variables. I could do a pca on my x
>>> variables getting eigen values and vectors then use predict pc1 (a
>>> newly created var), focusing only on the first eigenvector. However,
>>> if one does not have a y variable what is the score actually giving us
>>> for individual observations? What is it a predicted value for? Also,
>>> in theory I do not see why the sum of these observations must add to 0
>>> but they are. If someone understands this I would greatly appreciate
>>> an explanation. Thanks!
*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index